feat: add /v1/models API

jtsang4 · May 11, 2023 · 564b907 · 564b907
1 parent ae53760
commit 564b907
Show file tree

Hide file tree

Showing 7 changed files with 149 additions and 16 deletions.
diff --git a/.vscode/settings.json b/.vscode/settings.json
@@ -0,0 +1,7 @@
+{
+  "[python]": {
+    "editor.defaultFormatter": "ms-python.black-formatter"
+  },
+  "python.formatting.provider": "none",
+  "editor.formatOnSave": true
+}
diff --git a/claude_to_chatgpt/adapter.py b/claude_to_chatgpt/adapter.py
@@ -17,6 +17,7 @@
     "max_tokens": "length",
 }
 
+
 class ClaudeAdapter:
     def __init__(self, claude_base_url="https://api.anthropic.com"):
         self.claude_api_key = os.getenv("CLAUDE_API_KEY", None)
@@ -50,16 +51,16 @@ def openai_to_claude_params(self, openai_params):
             "max_tokens_to_sample": 9016,
         }
 
-        if (openai_params.get("max_tokens")):
+        if openai_params.get("max_tokens"):
             claude_params["max_tokens_to_sample"] = openai_params["max_tokens"]
 
-        if (openai_params.get("stop")):
+        if openai_params.get("stop"):
             claude_params["stop_sequences"] = openai_params.get("stop")
 
-        if (openai_params.get("temperature")):
+        if openai_params.get("temperature"):
             claude_params["temperature"] = openai_params.get("temperature")
-        
-        if (openai_params.get('stream')):
+
+        if openai_params.get("stream"):
             claude_params["stream"] = True
 
         return claude_params
@@ -79,10 +80,14 @@ def claude_to_chatgpt_response_stream(self, claude_response, prev_decoded_respon
                 {
                     "delta": {
                         "role": "assistant",
-                        "content": claude_response.get("completion", "").removeprefix(prev_decoded_response.get("completion", "")),
+                        "content": claude_response.get("completion", "").removeprefix(
+                            prev_decoded_response.get("completion", "")
+                        ),
                     },
                     "index": 0,
-                    "finish_reason": stop_reason_map[claude_response.get("stop_reason")] if claude_response.get("stop_reason") else None,
+                    "finish_reason": stop_reason_map[claude_response.get("stop_reason")]
+                    if claude_response.get("stop_reason")
+                    else None,
                 }
             ],
         }
@@ -107,7 +112,9 @@ def claude_to_chatgpt_response(self, claude_response):
                         "content": claude_response.get("completion", ""),
                     },
                     "index": 0,
-                    "finish_reason": stop_reason_map[claude_response.get("stop_reason")] if claude_response.get("stop_reason") else None,
+                    "finish_reason": stop_reason_map[claude_response.get("stop_reason")]
+                    if claude_response.get("stop_reason")
+                    else None,
                 }
             ],
         }
@@ -143,12 +150,20 @@ async def chat(self, request: Request):
                             try:
                                 decoded_line = json.loads(stripped_line)
                                 # yield decoded_line
-                                openai_response = self.claude_to_chatgpt_response_stream(decoded_line, prev_decoded_line)
+                                openai_response = (
+                                    self.claude_to_chatgpt_response_stream(
+                                        decoded_line, prev_decoded_line
+                                    )
+                                )
                                 prev_decoded_line = decoded_line
                                 yield openai_response
                             except json.JSONDecodeError as e:
-                                logger.debug(f"Error decoding JSON: {e}")  # Debug output
-                                logger.debug(f"Failed to decode line: {stripped_line}")  # Debug output
+                                logger.debug(
+                                    f"Error decoding JSON: {e}"
+                                )  # Debug output
+                                logger.debug(
+                                    f"Failed to decode line: {stripped_line}"
+                                )  # Debug output
             else:
                 claude_response = response.json()
                 openai_response = self.claude_to_chatgpt_response(claude_response)

diff --git a/claude_to_chatgpt/app.py b/claude_to_chatgpt/app.py
@@ -5,6 +5,7 @@
 import json
 import os
 from claude_to_chatgpt.logger import logger
+from claude_to_chatgpt.models import models_list
 
 CLAUDE_BASE_URL = os.getenv("CLAUDE_BASE_URL", "https://api.anthropic.com")
 LOG_LEVEL = os.getenv("LOG_LEVEL", "info")
@@ -25,26 +26,37 @@
     allow_headers=["*"],
 )
 
+
 @app.api_route(
     "/v1/chat/completions",
     methods=["POST", "OPTIONS"],
 )
 async def chat(request: Request):
     openai_params = await request.json()
     if openai_params.get("stream", False):
+
         async def generate():
             async for response in adapter.chat(request):
                 if response == "[DONE]":
                     yield "data: [DONE]"
                     break
                 yield f"data: {json.dumps(response)}\n\n"
+
         return StreamingResponse(generate(), media_type="text/event-stream")
     else:
         openai_response = None
-        async for response in adapter.chat(request):
-            openai_response = response
+        response = adapter.chat(request)
+        openai_response = await response.__anext__()
         return JSONResponse(content=openai_response)
 
+
+@app.route("/v1/models", methods=["GET"])
+async def models(request: Request):
+    # return a dict with key "object" and "data", "object" value is "list", "data" values is models list
+    return JSONResponse(content={"object": "list", "data": models_list})
+
+
 if __name__ == "__main__":
     import uvicorn
+
     uvicorn.run("app:app", host="0.0.0.0", port=PORT, log_level=LOG_LEVEL)
diff --git a/claude_to_chatgpt/logger.py b/claude_to_chatgpt/logger.py
@@ -1,3 +1,3 @@
 import logging
 
-logger = logging.getLogger("debug")
+logger = logging.getLogger("debug")
diff --git a/claude_to_chatgpt/models.py b/claude_to_chatgpt/models.py
@@ -0,0 +1,98 @@
+models_list = [
+    {
+        "id": "gpt-3.5-turbo",
+        "object": "model",
+        "created": 1677610602,
+        "owned_by": "openai",
+        "permission": [
+            {
+                "id": "modelperm-YO9wdQnaovI4GD1HLV59M0AV",
+                "object": "model_permission",
+                "created": 1683753011,
+                "allow_create_engine": False,
+                "allow_sampling": True,
+                "allow_logprobs": True,
+                "allow_search_indices": False,
+                "allow_view": True,
+                "allow_fine_tuning": False,
+                "organization": "*",
+                "group": None,
+                "is_blocking": False,
+            }
+        ],
+        "root": "gpt-3.5-turbo",
+        "parent": None,
+    },
+    {
+        "id": "gpt-3.5-turbo-0301",
+        "object": "model",
+        "created": 1677649963,
+        "owned_by": "openai",
+        "permission": [
+            {
+                "id": "modelperm-tsdKKNwiNtHfnKWWTkKChjoo",
+                "object": "model_permission",
+                "created": 1683753015,
+                "allow_create_engine": False,
+                "allow_sampling": True,
+                "allow_logprobs": True,
+                "allow_search_indices": False,
+                "allow_view": True,
+                "allow_fine_tuning": False,
+                "organization": "*",
+                "group": None,
+                "is_blocking": False,
+            }
+        ],
+        "root": "gpt-3.5-turbo-0301",
+        "parent": None,
+    },
+    {
+        "id": "gpt-4",
+        "object": "model",
+        "created": 1678604602,
+        "owned_by": "openai",
+        "permission": [
+            {
+                "id": "modelperm-nqKDpzYoZMlqbIltZojY48n9",
+                "object": "model_permission",
+                "created": 1683768705,
+                "allow_create_engine": False,
+                "allow_sampling": False,
+                "allow_logprobs": False,
+                "allow_search_indices": False,
+                "allow_view": False,
+                "allow_fine_tuning": False,
+                "organization": "*",
+                "group": None,
+                "is_blocking": False,
+            }
+        ],
+        "root": "gpt-4",
+        "parent": None,
+    },
+    {
+        "id": "gpt-4-0314",
+        "object": "model",
+        "created": 1678604601,
+        "owned_by": "openai",
+        "permission": [
+            {
+                "id": "modelperm-PGbNkIIZZLRipow1uFL0LCvV",
+                "object": "model_permission",
+                "created": 1683768678,
+                "allow_create_engine": False,
+                "allow_sampling": False,
+                "allow_logprobs": False,
+                "allow_search_indices": False,
+                "allow_view": False,
+                "allow_fine_tuning": False,
+                "organization": "*",
+                "group": None,
+                "is_blocking": False,
+            }
+        ],
+        "root": "gpt-4-0314",
+        "parent": None,
+    },
+]
diff --git a/claude_to_chatgpt/util.py b/claude_to_chatgpt/util.py
@@ -1,7 +1,8 @@
 import tiktoken
 
+
 def num_tokens_from_string(string: str, encoding_name: str = "cl100k_base") -> int:
     """Returns the number of tokens in a text string."""
     encoding = tiktoken.get_encoding(encoding_name)
     num_tokens = len(encoding.encode(string))
-    return num_tokens
+    return num_tokens
diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "claude-to-chatgpt"
-version = "0.1.0"
+version = "0.2.0"
 description = ""
 authors = ["jtsang4 <wtzeng1@gmail.com>"]
 readme = "README.md"